LagerNVS：面向全神经实时新视角合成的潜空间几何
LagerNVS: Latent Geometry for Fully Neural Real-time Novel View Synthesis
用显式三维重建预训练的编码器为前馈式新视角合成注入三维归纳偏置，配合轻量解码器即可实时渲染并达到 31.4 PSNR；该思路可直接迁移到任何需要三维先验的潜变量生成或重建管线，并可替换为扩散解码器做生成式外推。
原文依据 The encoder is initialized from a 3D reconstruction network pre-trained using explicit 3D supervision.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论近期工作表明，神经网络无需显式三维重建即可完成新视角合成等三维任务。即便如此，我们认为强三维归纳偏置在这类网络的设计中仍然有益。我们通过提出 LagerNVS 来证明这一点：这是一个基于「三维感知」潜变量特征的编码器-解码器新视角合成网络。编码器由使用显式三维监督预训练的三维重建网络初始化，并与一个轻量解码器配对，以光度损失进行端到端训练。LagerNVS 在确定性前馈新视角合成上达到最优水平（在 Re10k 上峰值信噪比达 31.4），在已知与未知相机参数两种设定下均如此，可实时渲染，能泛化到野外数据，并可与扩散解码器配对以实现生成式外推。代码、模型与示例见 szymanowiczs.github.io/lagernvs。
英文原文摘要 900 字符
Recent work has shown that neural networks can perform 3D tasks such as Novel View Synthesis (NVS) without explicit 3D reconstruction. Even so, we argue that strong 3D inductive biases are still helpful in the design of such networks. We show this point by introducing LagerNVS, an encoder-decoder neural network for NVS that builds on `3D-aware' latent features. The encoder is initialized from a 3D reconstruction network pre-trained using explicit 3D supervision. This is paired with a lightweight decoder, and trained end-to-end with photometric losses. LagerNVS achieves state-of-the-art deterministic feed-forward Novel View Synthesis (including 31.4 PSNR on Re10k), with and without known cameras, renders in real time, generalizes to in-the-wild data, and can be paired with a diffusion decoder for generative extrapolation. See szymanowiczs.github.io/lagernvs for code, models, and examples.